This revised competing continuation proposal is directed at obtaining improved molecular genetic markers for prognosis and risk assessment in colorectal cancer (CRCa). Our group reported previously that specific and generalized genetic alterations in primary CRCa are associated with distant metastases and poor prognosis. These alterations include allelic deletion on chromosome 18q (DCC gene), deletion on chromosome 17p (p53 gene), and high fractional allelic loss. Our specific aims are: Specific Aim #1: Evaluate the prognostic value of the candidate clonal molecular genetic alterations in a prospective study of colonic and rectal carcinomas with moderate risk of recurrence based on stage. We will assess deletion of the DCC gene, deletion of the p53 gene, and fractional allelic loss in a new consecutive series of curatively resected T3NO-lMO adenocarcinomas of the right colon, left colon, and rectum. We will continue long-term followup of the patients and analyze their survival using life table and hazard analysis. Specific Aim #2: Determine the association of the candidate molecular genetic alterations with pathological and cellular phenotypic features of the carcinomas: a. Pathological features including luminal and mural growth patterns, differentiation, vascular invasion, host inflammatory response, and metastasis. b. Altered total DNA content (aneuploidy) by flow cytometry and image analysis. c. Rapid tumor cell proliferation as assessed by immunohistochemistry for proliferating cell nuclear antigens (PCNA) with monoclonal antibodies Ki-67 and 19F4. d. Altered nuclear skeleton manifested by altered nuclear morphometry as assessed by image analysis of Feulgen preparations and PCNA immunohistochemistry sections. We will correlate the molecular genetic features with pathological characteristics, cellular phenotype, and patient characteristics including family history using uni- and multivariate analysis, and attempt to define mechanisms underlying biological behavior. We will use decision analysis to evaluate prognostic utility. Specific Aim #3: Define genetic markers associated with the susceptibility to colorectal cancer. Large multigenerational cancer family pedigrees (Hereditary Non polyposis Colorectal Cancer Syndrome) will be examined using parametric and nonparametric linkage techniques. The studies will focus on markers shown to exhibit high frequencies of allelic loss in sporadic colorectal cancers in the studies described in Specific Aim #1.